Search Results

The ARK EDDP Assay is an immunoassay intended for the qualitative and/or semiquantitative determination of EDDP in human urine at cutoff concentrations of 100 ng/mL and 300 ng/mL. The assay is intended for use in laboratories with automated clinical chemistry analyzers. This in vitro diagnostic device is for prescription use only.

The semiquantitative mode is for the purpose of (1) enabling laboratories to determine an appropriate dilution of the specimen for confirmation by a confirmatory method, such as Gas Chromatography/Mass Spectrometry (GC/MS) or Liquid Chromatography/tandem Mass Spectrometry (LC-MS/MS), or (2) permitting laboratories to establish quality control procedures.

The ARK EDDP Assay provides only a preliminary analytical test result. A more specific alternative chemical method must be used in order to obtain a confirmed positive analytical result. Gas Chromatography/Mass Spectrometry (GC/MS) or Liquid Chromatography/tandem Mass Spectrometry (LC-MS/MS) is the preferred confirmatory method. Clinical consideration and professional judgment should be exercised with any drug test result, particularly when the preliminary test result is positive.

Device Description

The ARK EDDP Assay is a homogeneous enzyme immunoassay technique used for the analysis of EDDP in human urine. The assay is based on competition between EDDP in the specimen and EDDP labeled with recombinant glucose-6-phosphate dehydrogenase (rG6PDH) for antibody binding sites. As the latter binds antibody, enzyme activity decreases. In the presence of EDDP from the specimen, enzyme activity increases and is directly related to the EDDP concentration. Active enzyme converts nicotinamide adenine dinucleotide (NAD) to NADH in the presence of glucose-6-phosphate (G6P), resulting in an absorbance change that is measured spectrophotometrically. Endogenous G6PDH does not interfere because the coenzyme NAD functions only with the bacterial enzyme used in the assay.

The ARK EDDP Assay consists of reagents R1 anti-EDDP rabbit antibody with substrate and R2 EDDP derivative labeled with bacterial recombinant G6PDH enzyme.

AI/ML Overview

Here's a breakdown of the acceptance criteria and study details for the ARK EDDP Assay, based on the provided FDA 510(k) summary:

This document describes a medical device, the ARK EDDP Assay, which is an immunoassay for the qualitative and/or semi-quantitative determination of EDDP (a methadone metabolite) in human urine. The study presented is a non-clinical performance evaluation, as indicated by "Brief Description of Nonclinical and Clinical Data" and the nature of the tests (Precision, Analytical Recovery, Analytical Specificity, Interference, etc.). No clinical study involving human patients or human readers in an MRMC setting is described.

1. A table of acceptance criteria and the reported device performance

The document doesn't explicitly state "acceptance criteria" as a single, consolidated table with specific pass/fail values for each performance characteristic. Instead, performance is evaluated against expected behaviors for an in vitro diagnostic (IVD) device of this type. The inferred acceptance criteria are that the device should demonstrate acceptable precision, accurate analytical recovery, appropriate cross-reactivity and interference profiles, and good agreement with a gold standard (GC/MS) in method comparison for qualitative and semi-quantitative results.

Performance Characteristic	Inferred Acceptance Criteria (Implicit)	Reported Device Performance and Discussion
Precision	Consistent qualitative and semi-quantitative results, particularly around the cutoff concentrations.	100 ng/mL Cutoff:
- Qualitative (100 ng/mL)	All samples ≤ 75 ng/mL should be negative; all samples ≥ 125 ng/mL should be positive. Samples at 100 ng/mL (cutoff) may show mixed results.	At 0.0, 25.0, 50.0, and 75.0 ng/mL, all 160 results were Negative. At 125.0, 150.0, 175.0, and 200.0 ng/mL, all 160 results were Positive. At the 100.0 ng/mL cutoff, 123 were Negative and 37 were Positive, which is expected for a cutoff concentration.
- Semiquantitative (100 ng/mL)	Mean measured concentrations should be close to theoretical values. Expected mixed qualitative results at cutoff.	Mean concentrations were close to theoretical (e.g., 98.1 ng/mL for 100 ng/mL theoretical). At 100.0 ng/mL cutoff, 114 were Negative and 46 were Positive, consistent with a cutoff.
- Qualitative (300 ng/mL)	All samples ≤ 225 ng/mL should be negative; all samples ≥ 375 ng/mL should be positive. Samples at 300 ng/mL (cutoff) may show mixed results.	At 0.0, 75.0, 150.0, and 225.0 ng/mL, all 160 results were Negative. At 375.0, 450.0, 525.0, and 600.0 ng/mL, all 160 results were Positive. At the 300.0 ng/mL cutoff, 57 were Negative and 103 were Positive, which is expected.
- Semiquantitative (300 ng/mL)	Mean measured concentrations should be close to theoretical values. Expected mixed qualitative results at cutoff.	Mean concentrations were close to theoretical (e.g., 298.8 ng/mL for 300 ng/mL theoretical). At 300.0 ng/mL cutoff, 85 were Negative and 75 were Positive, consistent with a cutoff.
Analytical Recovery	Measured concentration should be within an acceptable percentage range of the theoretical concentration.	Recovery percentages ranged from 94.6% to 107.9% across EDDP concentrations from 50.0 to 1000.0 ng/mL, indicating good analytical recovery.
Analytical Specificity	Minimal to no cross-reactivity with structurally related and unrelated compounds, ensuring specificity for EDDP.	Structurally Related Compounds: EDDP showed 100% cross-reactivity (as expected). Methadone and EMDP showed very low cross-reactivity (<0.005% to <0.03%). Several other structurally related compounds (Chlorpromazine, Diphenhydramine, Methylphenidate, Doxylamine) showed low cross-reactivity (<0.1% to <0.3%). This indicates good specificity for EDDP while acknowledging minor cross-reactivity with very high concentrations of certain related compounds.
Interference	Endogenous substances, specific gravity variations, and pH variations should not cause false positive or false negative results.	Structurally Unrelated Compounds: Numerous compounds were tested at high concentrations (e.g., 100,000 ng/mL, 500,000 ng/mL). For both 100 ng/mL and 300 ng/mL cutoffs, none of the listed substances yielded a false result (i.e., they were negative at -25% cutoff EDDP concentration and positive at +25% cutoff EDDP concentration), indicating no significant interference from these compounds even at very high concentrations.
		Endogenous Substances: High concentrations of common endogenous substances (e.g., Acetone, Ascorbic Acid, Bilirubin, Creatinine, Ethanol, Glucose, Hemoglobin, etc.) were tested. No interference was observed for either the 100 ng/mL or 300 ng/mL cutoff, meaning they did not cause false results.
		Specific Gravity: Urine samples across the physiologic range (1.002 to 1.030) did not show interference for either cutoff, meaning results remained correct (negative at -25% cutoff EDDP, positive at +25% cutoff EDDP).
		pH: Urine samples across the physiologic pH range (3.0 to 11.0) did not show interference for either cutoff, meaning results remained correct (negative at -25% cutoff EDDP, positive at +25% cutoff EDDP).
Method Comparison	High concordance with the GC/MS gold standard, particularly for samples well above or below the cutoff. Minor discrepancies near the cutoff are acceptable.	100 ng/mL Cutoff: 104 out of 109 samples showed perfect agreement with GC/MS (40 Low Negative, 60 High Positive). 5 samples were Near Cutoff Negative by GC/MS but Negative by immunoassay (expected). 4 samples were Near Cutoff Positive by GC/MS but Positive by immunoassay (expected). This shows high concordance.
		300 ng/mL Cutoff: 104 out of 109 samples showed perfect agreement with GC/MS (49 Low Negative, 52 High Positive). 4 samples were Near Cutoff Negative by GC/MS but Negative by immunoassay (expected). 3 samples were Near Cutoff Positive by GC/MS but Positive by immunoassay (expected). One discordant result: Sample 51 was Immunoassay Positive but 294 ng/mL by GC/MS (which is slightly below the 300 ng/mL cutoff). This is a single false positive near the cutoff, which can occur for immunoassay methods.
Calibration Curve Stability	Calibration should remain stable for a reasonable period.	A stored calibration curve was effective up to at least 15 days, which is an acceptable stability period for a laboratory assay.

2. Sample sizes used for the test set and the data provenance

Precision Studies:
- Target Concentrations: 9 levels for each cutoff (0.0 to 200.0 ng/mL for 100 ng/mL cutoff; 0.0 to 600.0 ng/mL for 300 ng/mL cutoff).
- Replicates: Quaduplicate twice a day for 20 days.
- Total Sample Size per Cutoff (for precision): 9 levels * 4 replicates * 2 (times/day) * 20 days = 1440 individual readings for the qualitative precision tables. (Note: The table summary states N=160 for each concentration level, implying 160 results for each of the 9 spiked levels.)
- Data Provenance: Drug-free, negative human urine (spiked with EDDP). The country of origin is not specified but is likely the U.S. as it's an FDA submission. This is a prospective study (laboratory testing).
Analytical Recovery:
- Target Concentrations: 12 levels (50.0 to 1000.0 ng/mL).
- Replicates: N=6 for mean concentration calculation.
- Total Sample Size: 12 levels * 6 replicates = 72 individual readings.
- Data Provenance: Drug-free, negative human urine (spiked with EDDP). Prospective laboratory testing.
Analytical Specificity (Cross-reactivity):
- Number of Compounds Tested: 7 structurally related compounds.
- Sample Size: Each compound tested at minimum concentration for positive result (or highest tested) at both cutoff levels. Implies multiple tests per compound, but specific N is not given per compound; typically, this would be triplicate.
- Data Provenance: Drug-free, negative human urine (spiked with compounds). Prospective laboratory testing.
Interference (Structurally Unrelated, Endogenous Substances, Specific Gravity, pH):
- Number of Compounds/Conditions Tested: Numerous (tens of compounds + SG and pH conditions).
- Sample Size: Each compound/condition tested at two EDDP levels (±25% of cutoff). Implies multiple tests per condition, but specific N is not given per condition; typically, this would be triplicate.
- Data Provenance: Drug-free, negative human urine (spiked with EDDP and interferent). Prospective laboratory testing.
Method Comparison:
- Sample Size: 109 unaltered clinical human urine specimens.
- Data Provenance: "unaltered clinical human urine specimens that are not individually identifiable." Country of origin not specified, but usually implies local (US) clinical samples. Described as "clinical human urine specimens," suggesting these were collected from actual patients, making this a retrospective evaluation of real-world samples.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

The ground truth for this diagnostic device is established by Gas Chromatography/Mass Spectrometry (GC/MS) or Liquid Chromatography/tandem Mass Spectrometry (LC-MS/MS), which are highly accurate analytical chemistry techniques considered the gold standard for drug confirmation in urine. These methods are performed in laboratories by trained laboratory professionals (analytical chemists, toxicologists). The document states the GC/MS confirmatory method was performed by a "licensed reference laboratory."

Number of experts: Not explicitly stated, as GC/MS analysis is a scientific measurement, not typically an "expert reader" interpretation process like image analysis. The "experts" are the qualified laboratory personnel operating and interpreting the GC/MS results.
Qualifications of experts: Implied to be qualified laboratory professionals with expertise in analytical chemistry and toxicology, as they work in a "licensed reference laboratory." Specific certifications or experience years are not given.

4. Adjudication method for the test set

Not applicable. The ground truth is established by a definitive analytical method (GC/MS). There is no "adjudication" in the sense of reconciling disagreements among multiple human readers for an image-based or qualitative assessment. The GC/MS provides a quantitative concentration, which is then used to determine positive/negative relative to the cutoffs. Any "discordance" (like the one noted in the method comparison for Sample 51) is a direct comparison between the immunoassay result and the GC/MS result, not a result of human reader discrepancy requiring adjudication.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

No, a multi-reader multi-case (MRMC) comparative effectiveness study was not done. This device is an in vitro diagnostic (IVD) assay designed to provide an automated analytical result (qualitative or semi-quantitative concentration of EDDP). It is not an imaging AI device that assists human readers in interpreting medical images. Therefore, the concept of "human readers improve with AI vs without AI assistance" is not applicable here.

6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done

Yes, the primary performance evaluation is standalone. The ARK EDDP Assay is an automated immunoassay performed on clinical chemistry analyzers. The results (qualitative positive/negative or semi-quantitative concentration) are generated by the assay system (reagents + analyzer) without human interpretive input for the result itself. The results are then reported to a clinician for medical interpretation in conjunction with other clinical information. The precision, analytical recovery, specificity, interference, and method comparison studies all represent the "algorithm only" or "device only" performance.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

The ground truth used is definitive analytical method (Gas Chromatography/Mass Spectrometry - GC/MS). For in vitro diagnostic assays measuring specific chemical compounds, GC/MS is considered the most accurate and reliable "truth" method.

8. The sample size for the training set

This document does not specify a separate "training set" sample size. For an immunoassay like the ARK EDDP Assay, the "training" (development and optimization) phase would involve numerous laboratory experiments to select reagents, optimize assay conditions, and establish initial performance characteristics. This is distinct from machine learning model training sets. The studies presented (precision, analytical recovery, specificity, interference, method comparison) are performance validation studies for the final device, not datasets used to train an algorithm.

9. How the ground truth for the training set was established

Not applicable in the context of an immunoassay. The "ground truth" for developing an immunoassay would be established by controlled spiking experiments using certified reference materials of the analyte (EDDP) in a known matrix (drug-free human urine), and confirming those concentrations using methods like GC/MS or LC-MS/MS, or by gravimetric/volumetric preparation. The goal is to ensure the assay accurately measures the target compound.

Ask a Question

Ask a specific question about this device

Page 1 of 1